A novel data-aided OFDM symbol timing synchronization method was proposed to reduce the computational cost and achieve more robust performance . The algorithm uses histograms to process the received signal's phase information for estimating symbol timing delay. In this paper, the histogrambased algorithm is extended further to cover MIMO-OFDM systems. In addition, detailed analysis is carried out to unveil the unique characteristics of the algorithm, in contrast to popular synchronization methods based on time-domain correlations. In particular, the algorithm converges on the mass center of a channel's power profile as the channel delay estimation. Good balance between performance and cost makes the algorithm a very attractive algorithm of choice for symbol timing synchronization.